Giulia Fanti [*]
[Full text of this Article in PDF is available at this link]
Introduction
Today, most central banks worldwide are exploring some form of central-bank digital currency (CBDC), a digital form of central bank money accessible to the public.1 There has been particular interest in cross-border CBDCs (also commonly called multi-CBDCs), which can be used to transfer assets from a CBDC ledger in one jurisdiction (typically one country) to another.2
Important open questions surround how to design multi-CBDCs. For example, how should the system be architected? How should data flow? How should transactions be processed and settled? How should the system be governed?
In general, these questions remain open. Part of the challenge is that multi-CBDCs must satisfy many desired properties, which can sometimes interfere with one another. In this Article, I discuss the tensions between three desired properties for cross-border CBDCs: security, privacy, and performance. I present a trilemma, which states that existing designs for multiCBDCs do not achieve all three desired properties. I then illustrate how existing common designs for multi-CBDCs fail to achieve all three properties. However, I also argue that the limitations of current implementations are not fundamental. I believe that with proper cooperation and collaboration between stakeholders, these technical challenges can and will be circumvented, enabling secure, private, and performant cross-border CBDC transactions.
In the remainder of the Article, I will assume that a cross-border CBDC would be built upon distributed ledger technology (DLT). A paper published by the Bank of International Settlements defines DLT as “the protocols and supporting infrastructure that allow computers in different locations to propose and validate transactions and update records in a synchronised way across a network.”3 DLT is a natural design choice for multi-CBDCs, in which there is no central trusted party. Indeed, DLT has been the technology of choice in many early pilot multi-CBDC programs,4 allowing independent domestic CBDC ledgers to be interlinked without requiring all CBDCs to interface on the same platform.
To my knowledge, every multi-CBDC pilot study to date has adopted an enterprise DLT solution. These enterprise DLT solutions are commercial software products that allow one or more organizations to maintain a DLT amongst themselves. For example, R3 has built a DLT platform called Corda, which has been used in several CBDC pilot studies.5 While these enterprise solutions are practical in many respects, they do not currently cover the full space of technical designs or properties one might envision in a multi-CBDC. Throughout the remainder of this Article, I will present concrete examples of how pilot projects have used enterprise DLT offerings, and how these products’ design choices and constraints affect the resulting multi-CBDC’s system properties.
[*] Assistant Professor of Electrical and Computer Engineering at Carnegie Mellon University. Ph.D. in Electrical Engineering and Computer Science from the University of California, Berkeley.
1 Central Bank Digital Currency Tracker, ATL. COUNCIL, https://www.atlanticcouncil.org/cbdctracker/ (last visited Oct. 17, 2022) [https://perma.cc/G92V-3ZPT] [hereinafter CBDC Tracker]; What is a Central Bank Digital Currency?, BD. OF GOVERNORS OF THE FED. RSRV.SYS. (Jan. 20, 2022), https://www.federalreserve.gov/faqs/what-is-a-central-bank-digital-currency.htm [https://perma.cc/DB5G-VZAU].
2 See CBDC Tracker, supra note 1.
3 Morten Bech & Rodney Garratt, Central Bank Cryptocurrencies, BIS Q.R.,Sept. 2017, at 55, 58. Note that DLT is a superset of blockchain technology; that is, blockchains are a form of DLT, but all DLT solutions are not blockchains.
4 E.g., BANK OF INT’L SETTLEMENTS, PROJECT JURA: CROSS-BORDER SETTLEMENT USING WHOLESALE CBDC 4 (2021), https://www.bis.org/publ/othp44.pdf [https://perma.cc/ZCN8- ZGGG] [hereinafter PROJECT JURA]; BANK OF INT’L SETTLEMENTS,PROJECT DUNBAR: INTERNATIONAL SETTLEMENTS USING MULTI-CBDCS 6 (2022), https://www.bis.org/publ/othp47.pdf [https://perma.cc/T2UT-WP3S] [hereinafter PROJECT DUNBAR]; BANK OF INT’L SETTLEMENTS, INTHANON-LIONROCK TO MBRIDGE: BUILDING A MULTI CBDC PLATFORM FOR INTERNATIONAL PAYMENTS 11 (2021), https://www.bis.org/publ/othp40.pdf [https://perma.cc/CF48-FNHC] [hereinafter INTHANONLIONROCK TO MBRIDGE]; BANK OF CAN. & MONETARY AUTH. OF SING., JASPER-UBIN DESIGN PAPER: ENABLING CROSS-BORDER HIGH VALUE TRANSFER USING DISTRIBUTED LEDGER TECHNOLOGIES 4 (2019), https://www.mas.gov.sg/-/media/Jasper-Ubin-DesignPaper.pdf [https://perma.cc/JGV4-7K2M] [hereinafter JASPER-UBIN DESIGN PAPER]; EUR. CENTRAL BANK & BANK OF JAPAN,SYNCHRONISED CROSS-BORDER PAYMENTS 1 (2019), https://www.ecb.europa.eu/paym/intro/publications/pdf/ecb.miptopical190604.en.pdf [https://perma.cc/H46Z-YZ9U].
5 PROJECT JURA, supra note 4, at 4; PROJECT DUNBAR, supra note 4, at 6; INTHANONLIONROCK TO MBRIDGE, supra note 4, at 6; JASPER-UBIN DESIGN PAPER, supra note 4, at 6.